ChattyPage
ChattyPagePage is an AI-based tool that allows users to interact with web-llm models directly within a browser environment. The tool employs a chat interface, grant...
Last verified:
What is ChattyPage?
ChattyPage is a private AI chat tool that runs large language models directly in the browser using WebGPU, so conversations are processed locally on your device rather than on a server. It is designed to feel familiar to users of modern chat interfaces while keeping the experience private and offline-capable after a model has been downloaded.
ChattyPage pricing
Pricing model: Free
The website presents Chatty as a live, open-source app rather than a paid subscription product. The GitHub documentation says you can try the app on the website for free, or run it locally from source with Node.js 18+ and npm, or via Docker. No paid tiers, usage limits, or premium plans are listed on the provided site materials.
ChattyPage pros
- Runs LLMs locally in the browser
- Keeps data on your device
- No server-side processing
- Works offline after model download
- Supports chat history
- Supports open-source models
- Works with Gemma models
- Works with Llama 2 and 3
- Works with Mistral models
- Responsive on supported phones
- Familiar ChatGPT-like interface
- Markdown rendering
- Code syntax highlighting
- Chat with uploaded files
- Custom memory instructions
- Export chats as JSON
- Export chats as Markdown
- Voice input support
- Regenerate responses
- Light and dark mode
ChattyPage cons
- Requires a WebGPU-capable browser
- Best supported in Chromium browsers
- Needs a capable GPU for efficient use
- 7B models need around 6GB VRAM
- 3B models need around 3GB VRAM
- Initial model download is required
- File embeddings are local but can be slower on smaller models
- Docker setup is not optimized for production
- Not all browsers have full support
- Multi-file embeddings are still on the roadmap
Frequently asked questions about ChattyPage
What is Chatty?
Chatty is a private AI chat application that runs large language models directly in your browser with WebGPU. The main idea is to give you a feature-rich AI chat experience while keeping processing local on your device instead of sending conversations to a server.
Does Chatty work offline?
Yes, once a model has been downloaded, Chatty can be used without an active internet connection. That offline behavior is part of its privacy-first design, since the model runs locally in the browser.
What browsers does Chatty support?
Chatty is designed for browsers with WebGPU support. The documentation says Chrome and Edge support WebGPU by default, and Firefox or Firefox Nightly may be enabled as well, but the browser must support the needed graphics features for the app to work well.
What models can I use with Chatty?
The project says it supports popular open-source models such as Gemma, Llama 2, Llama 3, and Mistral. Because the models run locally, performance depends on your hardware and the model size you choose.
Can Chatty read my files?
Yes, Chatty includes a file chat feature that lets you load files and ask questions about them. The documentation says PDF and other non-binary files are supported, and the processing stays local in your environment.
How does Chatty handle privacy?
Chatty is built around client-side inference, which means the AI models run on your own hardware in the browser. The project states that your data is processed only on your PC and that there is no server-side processing once the model is running locally.
Can I export my chats?
Yes, Chatty can export chat messages in either JSON or Markdown format. That makes it easier to save conversations, archive them, or move them into another workflow.
What hardware do I need?
The project notes that you need a GPU with enough memory to run models efficiently. It gives a rough guide of about 6GB of memory for 7B models and about 3GB for 3B models, so more capable hardware will deliver a smoother experience.
Can I run Chatty locally?
Yes, the repository includes local setup instructions using Node.js 18+ and npm, and it also provides Docker commands. That means you can try the web app or self-host it from source if you prefer a local installation.
Is Chatty open source?
Yes, Chatty is open source and published under the MIT license. The project also invites contributions and describes itself as built on open-source browser AI tooling.